Abstract
Background:
The lesbian, gay, bisexual, transgender, and queer/questioning (LGBTQ) community experiences health disparities. It is thus imperative that medical trainees receive training in the care of LGBTQ community. The objective of this study was to identify gaps in knowledge and comfort among medical school students in providing care for the LGBTQ community.
Methods:
An online survey was administered to medical students at 3 institutions in the United States from December 2020 to March 2021. Using a Likert scale, the survey assessed attitudes, comfort, and knowledge in providing care for the LGBTQ community. The survey included questions for each specific LGBTQ population. Results were quantified using descriptive and stratified analyses, and an exploratory factor analysis was used to calculate attitude summary measure (ASM) scores. A total knowledge score was calculated, with higher values indicating greater knowledge.
Results:
Among the 300 medical students who completed the survey, the majority were female (55.7%), White (54.7%), and heterosexual (64.3%). The majority of medical students felt comfortable (strongly agree/agree) participating in the care of lesbian (94.3%), gay (96.0%), and bisexual (96.3%) patients; this percentage dropped to 82.3% for non-binary and 71.3% for transgender patients. Only 27.0% of medical students reported confidence in their knowledge of health needs of transgender patients. LGBTQ self-identification, percent of core rotations completed in school, region of country, and friends and/or family who are part of the LGBTQ community were significantly associated with various ASM scores. Knowledge questions yielded high percentages of “neutral” responses, and medical students who identified as LGBTQ had significantly higher total knowledge scores.
Conclusions:
Overall, the surveyed medical students feel comfortable and willing to provide care for LGBTQ persons. But, there is limited knowledge about specific LGBTQ health needs. More education and training in the needs of transgender and non-binary patients, in particular, is indicated.
Introduction
Physical, mental, and sexual health disparities exist among the LGBTQ community. For example, the LGBTQ community face increased risks for cardiovascular disease, obesity, sexually transmitted infections, certain cancers, depression/suicide, and alcohol/substance abuse.1 -8 Despite increased risks, the LGBTQ community reports lower insurance coverage, increased discrimination in healthcare, and greater avoidance of healthcare due to fear of mistreatment.1,9 -15 Transgender patients are particularly less likely to seek healthcare because of discrimination and mistreatment.7,16,17
These findings compelled the American Medical Association and the American Association of Medical Colleges (AAMC) to seek enhanced education and research to address LGBTQ health disparities.18,19 Notwithstanding, LGBTQ-specific teaching in medical schools is limited. A survey study of 132 US/Canadian medical schools found the median time dedicated to LGBTQ teaching is 5 h across 4 years. 20 Further, literature on LGBTQ teaching is generally sparse, consisting primarily of single institution studies and focusing on LGBTQ as a single entity, as opposed to a heterogenous group with varied demographics, sexual orientations, gender identities, behaviors, and health needs.21 -26 Impact on overall learning within the medical school environment for LGBTQ students is largely unknown, as is how the medical school environment promotes or disengages LGBTQ students. A recent study found a correlation between lack of LGBTQ-centered teaching and LGBTQ students’ poor perception of their overall learning, highlighting that perceptions of the medical school environment may influence foundational learning and future practice patterns. 27
The current study measured knowledge and attitudes toward LGBTQ healthcare among medical students at 3 geographically distinct medical schools across the United States.
Methods
Study Population
The local Institutional Review Board (New York University, New York, NY) approved this study and was deemed exempt. Medical students from institutions in 3 US states were invited to participate in an anonymous survey-based research study from December 2020 to March 2021. As the local Institutional Review Board recognizes medical students as a protected class, the 3 institutions involved and responses were anonymized. Participation was voluntary and no incentives were provided. Study information and online consent preceded the Qualtrics survey, which 300 medical students completed. There were no inclusion or exclusion criteria. The only eligibility criterium was being enrolled into one of the 3 medical schools.
Survey Measures
The survey included 3 sections: Attitudes (31 questions), Knowledge (10 questions), and Demographics (13 questions) (Supplemental Digital Appendix 1). A glossary provided definitions of gender identity and sexual orientation.28,29 Survey items were modified from the National Oncologists Survey conducted by our group.30 -32 As participants were medical students, wording was changed to “participate in the care of” versus “provide care for” patients. Some questions were asked for each specific LGBTQ population. Attitude and knowledge items used a seven-point Likert scale: strongly agree (1), agree (2), somewhat agree (3), neutral (4), somewhat disagree (5), disagree (6), and strongly disagree (7). In the analysis, this was reversed with higher numbers corresponding with greater agreeability.
Statistical Analysis
Descriptive statistics were used to quantify responses, with frequencies and percentages for categorical variables and median and interquartile range (IQR) for continuous variables.
Stratified analyses assessed potential differences by pre-specified subgroups, including friends and/or family who are part of the LGBTQ community, LGBTQ self-identification, percent of core rotations completed, race/ethnicity, religion, and region of country. Region of country indicated location of each medical school and was defined using the US Census Bureau divisions: Middle Atlantic, MA (New Jersey, New York, Pennsylvania), South Atlantic, SA (Delaware, Florida, Georgia, Maryland, North/South Carolina, Virginia, West Virginia), and West South Central, WSC (Arkansas, Louisiana, Oklahoma, Texas). Kruskall-Wallis tests assessed relationships between survey items and respondent characteristics, with subsequent pairwise comparisons with a Bonferroni correction for multiple comparisons. Two moderation analyses examined the main effects of knowledge and comfort in treating, as well as their interaction on education and practice and comfort of inquiry using ordinary least-squares regression methods with the Hayes process macro. 33
An exploratory factor analysis using principal axis factoring and varimax rotation was conducted to determine factor structure of attitude items. Two items, “confidence in knowledge of health needs of transgender patients” and “importance of knowing the sexual orientation of patients to provide the best care,” did not meet factor loading criteria of ≥0.40 and were removed. The final exploratory factor analysis indicated 4 factors: comfort in treating (items 1-5, Table 2), education and practice (items 7, 9-12, Table 2), comfort of inquiry (items 13-16, Table 2;), and comfort discussing health needs (items 17-30, Table 2) (Further details in Supplemental Table S1). Within each factor, an attitude summary measure (ASM) score was calculated as the average of items. SPSS 24.0 (IBM, Chicago, Illinois) was used for all analyses.
Results
Demographics
The majority of the 300 students identified as female (167, 55.7%), non-Hispanic White (164, 54.7%), and heterosexual (193, 64.3%) (Table 1). There were 104 (34.7%) students who self-identified as LGBTQ and 286 (95.3%) had friends and/or family who are part of the LGBTQ community. Half the students (151, 50.3%) had not completed any core rotations, indicating they were in the preclinical phase of medical school; 44 (14.7%) had completed ≤50% and 105 (35.0%) had completed >50% of core rotations.
Demographic Characteristics of Medical Student Respondents (N = 300).
Abbreviations: IQR, interquartile range; LGBTQ, lesbian, gay, bisexual, transgender, queer/questioning.
One school did not receive the political affiliation item.
Attitude Measures
The majority of students were comfortable (strongly agree/agree) participating in the care of patients identifying as lesbian (283, 94.3%), gay (288, 96.0%), and bisexual (289, 96.3%). This percentage dropped to 82.3% for non-binary patients and 71.3% for transgender patients. Only 81 (27.0%) reported confidence in their knowledge of transgender patients’ healthcare. The majority of students agreed with statements regarding comfort asking patients’ name (283, 94.3%), partners (243, 81.0%), sexual behavior (228, 76.0%), pronouns (227, 75.7%), and sexual orientation (226, 75.3%). Fewer (192, 64.0%) agreed it is important to know patients’ sexual orientation to provide quality care, but 264 (88.0%) agreed it is important to know pronouns (Table 2).
Participant Responses to Select Attitude Items (N = 300).
Abbreviation: LGBTQ, lesbian, gay, bisexual, transgender and queer/questioning.
When asked about comfort discussing preventative, sexual, reproductive, and gender-affirming health needs, a higher proportion of students agreed with the respective statements for lesbian, gay and bisexual patients than for transgender and non-binary patients (Supplemental Table S2).
With respect to education, 269 (89.7%) agreed there should be mandatory education on LGBTQ health in medical school, and 272 (90.7%) would participate in such training (Table 2). ASM subscale medians indicated agreeability with the respective items in each subscale (Table 3).
Stratified Analysis of Attitudes Toward LGBTQ Health Among Surveyed Medical Students (N = 300).
Abbreviations: IQR, interquartile range; LGBTQ, lesbian, gay, bisexual, transgender, queer/questioning.
Knowledge Measures
There were varied responses to the knowledge questions. The majority of students correctly indicated that HPV-associated cervical dysplasia is not only found in cisgender women having heterosexual intercourse (253, 84.3%) and that LGBTQ persons avoid accessing healthcare (266, 88.7%). There were high percentages of “neutral” responses (39.0%-71.7%) for the remaining knowledge questions (Table 4). The median (IQR) total knowledge score was 4.9 (0.9) (Table 5).
Participant Responses to Knowledge Items (N = 300).
Abbreviations: LGBTQ, lesbian, gay, bisexual, transgender and queer/questioning; LGB, lesbian, gay, and bisexual; HPV, human papillomavirus; STI, sexual transmitted infection; HT, hormone therapy.
Stratified Analysis of Knowledge of LGBTQ Health Among Surveyed Medical Students (N = 300).
Abbreviations: IQR, interquartile range; LGBTQ, lesbian, gay, bisexual, transgender, queer/questioning; HPV, human papillomavirus; LGB, lesbian gay bisexual; STI, sexually transmitted infection; HT, hormone therapy.
Stratified Analyses
For the stratified analyses, distributions were not uniformly similar and comparisons of mean ranks are reported (Tables 3 and 5). Having friends and/or family who are part of the LGBTQ community and self-identifying as LGBTQ were associated with higher agreement with multiple attitude items (Table 3). LGBTQ self-identification was significantly associated with several knowledge items, but having friends and/or family who are part of the LGBTQ community was not (Table 5). Identifying as non-religious was associated with greater agreement with mandatory education on LGBTQ health (146.0vs 133.0, P = .01), willingness to participate in training if offered (147.0vs 132.0, P = .002), and willingness to be listed as an LGBTQ-friendly provider than identifying as religious (146.9vs 132.1, P = .001, Supplemental Table S3). No statistically significant associations were found for knowledge items based on religion (Supplemental Table S4).
Region of the country was associated with several attitude and knowledge items and adjusted P-values are presented for the pairwise comparisons with a Bonferroni correction (Tables 3 and 5). Students in WSC had lower agreement with mandatory education on LGBTQ health than students in MA and SA (136.8vs 164.0, P < .001 and 136.8 vs 154.9, P = .01, respectively) and with willingness to participate in LGBTQ health training if offered (140.5vs 158.9, P = .01 and 140.5 vs 154.8, P = .04, respectively). Likewise, students in WSC had lower agreement with willingness to be listed as an LGBTQ-friendly provider than students in SA (141.1vs 157.1, P = .01), but not MA. However, when asked about working with transgender patients, students in WSC indicated higher comfort discussing preventative health needs than students in MA (166.2vs 139.6, P = .04) and SA (166.2vs 140.6, P = .03), higher comfort discussing reproductive health needs than students in MA (171.2vs 138.5, P = .01) and SA (171.2vs 137.6, P = .004), and higher comfort discussing sexual health needs than students in MA (166.5vs 136.7, P = .02). Finally, students in MA indicated higher comfort caring for transgender patients than students in SA (170.8vs 133.2, P = .001). Students in WSC had lower agreement that LGBT patients avoid accessing healthcare than students in MA (136.8vs 155.3, P = .01) and students in SA (136.8vs 153.4, P = .01), that LGB cisgender women have worse birth outcomes than students in SA (132.3vs 164.7, P = .01), and that transgender youth are interested in fertility preservation prior to hormone replacement therapy than students in MA (134.6vs 166.2, P = .01).
Interestingly, students with ≤50% completed core clerkships indicated lower agreement with multiple attitude items than preclinical students and students with >50% completed core clerkships including lower comfort participating in the care of transgender patients (120.9vs 158.5, P = .004 and 120.9 vs 151.4, P = .04, respectively), lower comfort asking patients’ pronouns (117.3vs 159.4, P < .001 and 117.3 vs 151.5, P = .01, respectively), and lower agreement with willingness to be listed as an LGBTQ-friendly provider (131.3vs 153.7, P = .004 and 131.3 vs 153.9, P = .01, respectively). Students with ≤50% completed core clerkships indicated lower agreement than preclinical students with the importance of knowing patients’ pronouns (130.1vs 156.8, P = .004), mandatory education on LGBTQ health needs (134.5vs 155.2, P = .03), and willingness to participate in LGBTQ health training (133.4vs 156.7, P = .01). Progression through medical school was associated with greater knowledge on 2 items. Students with >50% completed core clerkships indicated lower agreement than preclinical students that HPV-associated dysplasia is only found in cisgender women (137.3vs 154.3, P = .03), and students with ≤50% completed core clerkships indicated higher agreement than preclinical students that LGBT individuals have high prevalence of substance use (168.6 vs 136.8, P = .04).
Comparing students with and without friends and/or family who are part of the LGBTQ community, those who did had higher median scores for the comfort in treating and education and practice ASM subscales [6.8 (1.0) vs 5.9 (2.7), P = .02 and 6.6 (0.8) vs 5.7 (3.5), P < .001, respectively]. LGBTQ self-identification was associated with higher scores for all subscales (P < .05). Students in SA had lower scores for the discuss health needs subscale than students in WSC [5.4 (1.9) vs 5.9 (1.8), P = .02). Preclinical students had higher scores for comfort in treating and education and practice subscales than students with ≤50% completed core rotations [6.8 (1.0) vs 6.4 (1.4), P = .01 and 6.8 (0.6) vs 6.3 (1.4), P = .03, respectively, Table 3). LGBTQ self-identification was associated with higher total knowledge scores (P < .001, Table 5).
Moderation Analyses
Two moderation analyses examined associations of knowledge, comfort in treating and their interaction (knowledge × comfort in treating) with the education and practice and comfort of inquiry ASM subscales. The model for education and practice was significant (P < .01, R 2 = .26). Greater knowledge and greater comfort in treating patients of different LGBTQ identities was associated with higher education and practice scores (b = 3.5, P < .01 and b = 2.9, P < .01, respectively). The interaction term was also associated with higher education and practice scores (b = −0.5, P < .01), indicating that comfort in treating moderates the relationship between total knowledge and education and practice. The model for comfort of inquiry was not significant (P = .6).
Discussion
This study assessed attitudes, comfort levels, knowledge, and desire for education/training, among medical students at 3 geographically distinct US medical schools, regarding the care of the LGBTQ community. Overall, this study revealed that the majority of students felt comfortable participating in the care of lesbian, gay, and bisexual patients, but the percentage decreased for transgender and non-binary patients. When asked specifically about preventative, sexual, reproductive, and gender-affirming health needs, fewer students reported comfort in discussing these issues with any of the patient identities, and the least with transgender patients. There was high agreement for mandatory education on LGBTQ health needs in medical school, as well as interest in participating in training if provided.
Overall, attitudes and comfort toward the LGBTQ community were positive among medical students. Using similar metrics, prior studies from our group conducted with oncologists and advanced practice professionals (APPs) also found a generally positive attitude toward LGBTQ persons.30,31 Although analyses were not conducted comparing these groups with students, it is worth noting that students indicated higher comfort with LGBTQ persons than other healthcare professionals.30,34,35 An important difference to note from our previous work is that this study separated the identities within the LGBTQ community into different questions, providing a more detailed assessment of respondents’ comfort levels caring for subpopulations.30 -32 Fewer students expressed comfort for all questions involving transgender and non-binary patients. Although most students agreed with general statements concerning their comfort in treating LGBTQ persons, when asked specifically about comfort in discussing preventative, sexual, reproductive, and gender affirming health needs, fewer students agreed with the statements. Particularly with transgender and non-binary patients, several questions yielded <50% of respondents indicating comfort (Supplemental Table S2). These findings are consistent with the literature, which reports a lack of teaching about transgender health.25,36 -38 These findings support the need for greater training in medical school on LGBTQ health, with a specific focus on addressing subpopulations including transgender and non-binary healthcare.
Attitudes were influenced by respondent characteristics. Similar to previous studies within the general US population, individuals with friends and/or family who are part of the LGBTQ community, and those who self-identified as LGBTQ, had more positive attitudes.30 -32,39 -42 LGBTQ self-identification was associated with higher scores for all ASM subscales, and having friends and/or family who are part of the LGBTQ community was associated with higher scores in the comfort in treating and education and practice subscales. Compared with students without friends and/or family who are part of the LGBTQ community, those with them had greater agreement that it is important to know patients’ sexual orientation and pronouns to provide the best care, and had greater comfort asking patients’ pronouns. Likewise, LGBTQ self-identification was associated with greater comfort asking patients’ pronouns. This stratification in awareness of the relevance and comfort in collecting sexual orientation and gender identity (SOGI) information is important, as it is recommended to collect and document SOGI as a component of addressing LGBTQ health disparities3,43,44 However it is important to note that in states within the United States, as well other countries, disclosing LGBTQ status may not be safe.45 -48 Prior research indicates most patients are willing to discuss SOGI with healthcare providers and consider these topics important.35,49 -52 However, healthcare workers’ discomfort with seeking this information prevents its collection and documentation.35,53 A recent study found students stopped asking SOGI questions as they progressed through medical school, citing lack of relevance to patient encounters, attending physicians’ negative influence, and reliance on patients’ records. 24 Although there were not consistent differences between stages of medical school regarding SOGI information in this study, the overall percentage of students who agreed it is important for patient care was lower than desired. This was particularly true for sexual orientation, where only 64% of students agreed it is important to know. As medical students are the future generations of physicians, it is imperative to emphasize the importance of knowing and understanding SOGI, and to foster an environment where all healthcare team members routinely ask for and document this information.
Further, students’ overall knowledge regarding LGBTQ health varied, with a range of 5.8%-88.7% correctly answering individual items. Moreover, responses often fell in the neutral range. As noted previously, later stage in medical school was associated with greater agreement with 2 knowledge items, but this pattern was not significant for total knowledge scores. The overall low knowledge level in this study is consistent with published findings from our prior surveys of oncologists and APPs and those of others.30 -32,54 -57 And it is consistent with a recent study of British medical students at a single institution, which found deficits in students’ knowledge of specific LGBTQ health problems, including dissatisfaction with health services, avoidance of cervical cancer screening, and increased alcohol/drug use. 21 Medical students’ low knowledge level is not surprising given the lack of curricular emphasis on LGBTQ health. A 2011 survey of 132 US/Canadian medical schools identified only one-third with LGBTQ health training. 20 A 2017-2018 AAMC report found 76% of participating schools included LGBTQ health themes, and that 50% of those reported only 3 or fewer relevant activities. 58 Despite some advances in the number of schools reporting LGBTQ health training, these results further highlight medical students’ lack of understanding of LGBTQ health issues, which is pervasive across institutions.54 -57 Although there were significant differences for 3 of the knowledge items across regions of the country (as described above), no clear pattern emerged and there was no significant difference in total knowledge scores between geographic regions. Lack of knowledge about LGBTQ health can cause graduating medical students to feel unprepared to treat LGBTQ person and can lead to reduced quality of care.24,59 -64 This study also revealed an increase in positive attitudes toward LGBTQ persons in medical students compared to prior analyses with practicing physicians and APPs, but not an increase in knowledge. Together, these studies emphasize the need for continued efforts in incorporating LGBTQ health into medical education, training, and licensing.63,64
Potential limitations should be acknowledged. Demographic characteristics between responders and non-responders could not be assessed as the survey was anonymous, and self-selection bias may have impacted findings since participation was voluntary. The cross-sectional study design and the large number of statistical tests limit the ability to draw causal conclusions on the observed associations; future prospective studies are needed. Additionally, the responding medical students may have felt social pressure not to indicate negative attitudes. Due to the opt-in nature of the study, results may be biased as nearly 36% of students self-identified as LGBTQ.
Despite limitations, there are several strengths to this study. Prior research has primarily focused on single institutions, and it has been noted that attitudes may vary across regions of the country; our study included 3 institutions in distinct regions.55,60,65,66 Attitude items in this study were separated into respective identities within the LGBTQ community, when possible, which recognizes the diversity within the community and their unique health issues. Finally, although slightly different from previous studies in its factors, this is the third study in which our group assessed ASM subscales related to dimensions of LGBTQ-related attitudes.30,31 Future studies are needed to test validity and reliability, but this novel scale has the potential to assess progress in LGBTQ-related attitudes in research, clinical settings, and education.
In a geographically diverse cohort of US medical students, there was limited knowledge about LGBTQ health needs; however, participating students revealed high comfort in interacting with this community and interest in further training on unique health needs of the LGBTQ community. Future research needs to determine effective strategies to consistently add comprehensive LGBTQ teaching to medical curriculum.
Supplemental Material
sj-docx-1-jpc-10.1177_21501319231186729 – Supplemental material for Assessing Medical Students’ Attitudes and Knowledge Regarding LGBTQ Health Needs Across the United States
Supplemental material, sj-docx-1-jpc-10.1177_21501319231186729 for Assessing Medical Students’ Attitudes and Knowledge Regarding LGBTQ Health Needs Across the United States by Hannah C. Karpel, Amani Sampson, Mia Charifson, Lydia A. Fein, Devin Murphy, Megan Sutter, Christina L. Tamargo, Gwendolyn P. Quinn and Matthew B. Schabath in Journal of Primary Care & Community Health
Footnotes
Acknowledgements
None
Prior Presentations
This was presented as a poster presentation at the American Society of Clinical Oncology (ASCO) Annual Meeting, Chicago, Illinois, June 2022.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: MB Schabath reports research funding from Bristol Myers Squibb and is a consultant for BMS. GP Quinn reports honoraria (FLO Health) and speakers’ bureau (CME Outfitters).
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
